# app.py from flask import Flask, request, jsonify, render_template_string import os import requests import json import logging from typing import Dict, Any, List import time app = Flask(__name__) logging.basicConfig(level=logging.INFO, format='%(asctime)s - %(levelname)s - %(message)s') # Configuration OLLAMA_BASE_URL = os.getenv('OLLAMA_BASE_URL', 'https://huggingface.co/spaces/tommytracx/ollama-api') ALLOWED_MODELS = os.getenv('ALLOWED_MODELS', 'llama2,llama2:13b,llama2:70b,codellama,neural-chat,gemma-3-270m').split(',') MAX_TOKENS = int(os.getenv('MAX_TOKENS', '2048')) TEMPERATURE = float(os.getenv('TEMPERATURE', '0.7')) class OllamaManager: def __init__(self, base_url: str): self.base_url = base_url.rstrip('/') self.available_models = ALLOWED_MODELS # Initialize with allowed models self.refresh_models() def refresh_models(self) -> None: """Refresh the list of available models from Ollama API, falling back to allowed models.""" try: response = requests.get(f"{self.base_url}/api/tags", timeout=10) response.raise_for_status() data = response.json() models = [model['name'] for model in data.get('models', [])] # Filter models to only include those in ALLOWED_MODELS self.available_models = [model for model in models if model in ALLOWED_MODELS] if not self.available_models: self.available_models = ALLOWED_MODELS logging.warning("No allowed models found in API response, using ALLOWED_MODELS") logging.info(f"Available models: {self.available_models}") except Exception as e: logging.error(f"Error refreshing models: {e}") self.available_models = ALLOWED_MODELS def list_models(self) -> List[str]: """Return the list of available models.""" return self.available_models def generate(self, model_name: str, prompt: str, **kwargs) -> Dict[str, Any]: """Generate text using a model.""" if model_name not in self.available_models: return {"status": "error", "message": f"Model {model_name} not available"} try: payload = { "model": model_name, "prompt": prompt, "stream": False, **kwargs } response = requests.post(f"{self.base_url}/api/generate", json=payload, timeout=120) response.raise_for_status() data = response.json() return { "status": "success", "response": data.get('response', ''), "model": model_name, "usage": data.get('usage', {}) } except Exception as e: logging.error(f"Error generating response: {e}") return {"status": "error", "message": str(e)} def health_check(self) -> Dict[str, Any]: """Check the health of the Ollama API.""" try: response = requests.get(f"{self.base_url}/api/tags", timeout=10) response.raise_for_status() return {"status": "healthy", "available_models": len(self.available_models)} except Exception as e: logging.error(f"Health check failed: {e}") return {"status": "unhealthy", "error": str(e)} # Initialize Ollama manager ollama_manager = OllamaManager(OLLAMA_BASE_URL) # HTML template for the chat interface HTML_TEMPLATE = '''
Chat with your local Ollama models through Hugging Face Spaces